The overall goal of this research career development award is to promote the development of Dr. Ron Keren as an independent patient-oriented clinical researcher. To achieve this goal, Dr. Keren has framed a comprehensive program that combines a preceptorship with J. Sanford Schwartz, M.D., formal didactic training in health services research and clinical epidemiology, and performance of original research. This combination of didactic and practical research experiences will allow Dr. Keren to expand and strengthen the set of skills he acquired in his fellowship training. Dr. Keren hopes to use this award to establish himself as an academic leader in pediatric outcomes and health services research, with an expertise in predictive modeling and economic evaluation. Dr. Keren's interest is in technology assessment and public health policy as they relate to newborn screening and disease prevention in general, and to the prediction and prevention of severe neonatal hyperbilirubinemia and kernicterus in particular. The specific aims of the research application are three-fold. First, he will develop a simple and practical clinical prediction rule that pediatricians and neonatologists can use to predict term and near-term newborns' risk of developing severe neonatal hyperbilirubinemia after discharge from the hospital. Second, the clinical prediction rule will be validated in an independent sample of ethnically and geographically diverse newborns, thus demonstrating its generalizability. And finally, a cost-effectiveness model will be developed to simulate the economic costs and clinical harms and benefits of using alternative prediction models in the prevention of extreme hyperbilirubinemia and kernicterus. Future research projects and grant submissions will be aimed at evaluating the effectiveness and cost-effectiveness of these prediction models when coupled with risk-specific preventive services in actual nursery settings. Current preventive services consist of closer out-patient follow-up for infants designated as high risk for the development of severe neonatal hyperbilirubinemia after discharge. However, the imminent availability of drugs, such as tin-mesoporphyrin, that can prevent hyperbilirubinemia in high-risk infants makes the application to develop valid prediction rules to risk stratify infants particularly timely, relevant, and important.